TY  - GEN
ID  - heidok36819
N2  - This dissertation aims to illustrate the potential of machine learning models in political psychology research. Drawing on previous theoretical and empirical work from related disciplines, I outline three shifts in how research in political psychology can be conducted. In detail, I explain the benefits of focusing on prediction, embracing more complexity in the modeling phase and using large datasets and novel data sources ? and how the usage of machine learning models can support these shifts. Representing the main contribution of this publication-based dissertation, I then present three original studies in which I investigated belief in conspiracy theories, individual sustainable behavior and political attitudes and behavior in online social media networks using different machine learning models. I concurrently discuss how using these models in my studies helped to gain deep insights into the psychological mechanisms underlying political cognitions and behavior. Based on my original studies and insights from other previous studies, I outline potential future directions for political psychological research. I discuss the advantages and limitations of machine learning models, important precautions for their application, strategies to increase their future usage and approaches for integrating them into current research practice.
A1  - Brandenstein, Nils
CY  - Heidelberg
UR  - https://archiv.ub.uni-heidelberg.de/volltextserver/36819/
AV  - public
Y1  - 2025///
TI  - Paving New Ways - The Case for Machine Learning in Political Psychology
ER  -